diff options
| author | Michael Niedermayer <michaelni@gmx.at> | 2013-06-02 22:18:29 +0200 |
|---|---|---|
| committer | Michael Niedermayer <michaelni@gmx.at> | 2013-06-03 13:18:40 +0200 |
| commit | 96b71a6ec509023c892aaf9e7597ed0accbafb32 (patch) | |
| tree | 50edffb297c53df9404c37950d64419817078da8 /libavcodec/jpeg2000dwt.c | |
| parent | 6507d86f07b366b6ccc8102b0c638347c39bf088 (diff) | |
| download | ffmpeg-96b71a6ec509023c892aaf9e7597ed0accbafb32.tar.gz | |
avcodec/jpeg2000dwt: merge rescaling with interleave in 9/7 float IDWT
10% faster
Reviewed-by: Nicolas BERTRAND <nicoinattendu@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Diffstat (limited to 'libavcodec/jpeg2000dwt.c')
| -rw-r--r-- | libavcodec/jpeg2000dwt.c | 15 |
1 files changed, 4 insertions, 11 deletions
diff --git a/libavcodec/jpeg2000dwt.c b/libavcodec/jpeg2000dwt.c index 7e82a56f19..13712d9703 100644 --- a/libavcodec/jpeg2000dwt.c +++ b/libavcodec/jpeg2000dwt.c @@ -359,13 +359,6 @@ static void sr_1d97_float(float *p, int i0, int i1) extend97_float(p, i0, i1); - /*step 1*/ - for (i = i0 / 2 - 1; i < i1 / 2 + 2; i++) - p[2 * i] *= F_LFTG_K; - /* step 2*/ - for (i = i0 / 2 - 2; i < i1 / 2 + 2; i++) - p[2 * i + 1] *= F_LFTG_X; - /* step 3*/ for (i = i0 / 2 - 1; i < i1 / 2 + 2; i++) p[2 * i] -= F_LFTG_DELTA * (p[2 * i - 1] + p[2 * i + 1]); /* step 4 */ @@ -401,9 +394,9 @@ static void dwt_decode97_float(DWTContext *s, float *t) int i, j = 0; // copy with interleaving for (i = mh; i < lh; i += 2, j++) - l[i] = data[w * lp + j]; + l[i] = data[w * lp + j] * F_LFTG_K; for (i = 1 - mh; i < lh; i += 2, j++) - l[i] = data[w * lp + j]; + l[i] = data[w * lp + j] * F_LFTG_X; sr_1d97_float(line, mh, mh + lh); @@ -417,9 +410,9 @@ static void dwt_decode97_float(DWTContext *s, float *t) int i, j = 0; // copy with interleaving for (i = mv; i < lv; i += 2, j++) - l[i] = data[w * j + lp]; + l[i] = data[w * j + lp] * F_LFTG_K; for (i = 1 - mv; i < lv; i += 2, j++) - l[i] = data[w * j + lp]; + l[i] = data[w * j + lp] * F_LFTG_X; sr_1d97_float(line, mv, mv + lv); |